[Tone Depth]

Loudspeaker Enclosure Modification – Part 1:  Background

I purchased used Tangent RS2 monitor speakers (see the first photo) in 1983, attracted by their imaging and sound quality. At normal listening levels, the midrange has always been enjoyable. Eventually, as time passed I noticed some design shortcomings in the RS2s, a midbass peak to rolling off maybe down 3 dB at about 55Hz, and highs that seemingly weren’t as revealing as more current technology tweeters. I don’t currently have the budget or floor space to replace the RS2s with floor standers, as both the TV and the stereo system are located in our family room and converting it from a multipurpose room to a dedicated listening room isn’t an option.



The speakers are a basic two-way sealed design (see the second photo), the crossover between tweeter and mid-bass driver reported at 3,000Hz, and with a typical British design influence. I’ve read references of theories that semi-resonant enclosures were thought to contribute special sound qualities to the speaker’s output. The enclosure is constructed from ½” MDF, 14” H x 10” W x 11” D, with teak veneer. I was surprised to discover only two inch acoustic open cell polyurethane foam lining the inside of the enclosure, until I realized this was a typical design/construction practice when the speakers were built. The Audax 8” nominal woofer has a 6.5” diameter Bextrene cone. The KEF T-27 melinex-domed tweeter was well respected in its day, and also used in the Rogers LS3/5A and many other popular speakers. Neither of these drivers is currently in production.

[Tone Depth]

Loudspeaker Enclosure Modification – Part 2:  Evaluation & Analysis

I noticed when I turned up the volume, that I could sense some distortion and listening fatigue. I also could feel vibrations with my fingertips while lightly touching the sides of the enclosures, especially when I turned up the volume. Since the surface area of the enclosure top, sides and rear panels (448 sq in) exceeds the surface area of the mid/woofer cone by 13.5 times, relatively low amplitude enclosure panel vibrations can create audible degradation of the sound quality. I often wondered how to more fully realize the full potential of the speaker performance. In seeking an answer to this question, I tried mass loading the speakers by adding weights on the top of the enclosures, as recommended by Frank Van Alstine of Audio by Van Alstine. I heard an audible improvement in the sound; however, I wasn’t satisfied with the aesthetic appearance of added mass sitting on the speaker enclosures.

[Tone Depth]

Loudspeaker Enclosure Modification – Part 3:  Alternatives

I read a white paper by Albert Von Schweikert of Von Schweikert Audio (VSA), in which he describes the patented triple wall laminate design http://www.audiocircle.com/index.php?topic=70291.0 he developed for his line of virtual reality (VR) loudspeakers, a high performance method to reduce audible coloration from cabinet panel resonances (see the third photo). The enclosures are constructed with ¾” resin-impregnated and veneered MDF with a medium Q factor, ¾” synthetic stone with a very high Q factor, and ¾” felt with a very low Q factor, with 5 mm layers of visco-elastic constrained damping compound between the three laminates to minimize enclosure wall resonances. Lining the triple laminate enclosure interior walls is graduated-density Dacron Polyester batting.

[Tone Depth]

Loudspeaker Enclosure Modification – Part 4:  Design Selection

I decided to adapt and scale the VSA triple laminate design using opposing Q factors to my requirements, since I would be modifying my existing veneered MDF enclosures (with a medium Q factor), and selected ½” Richlite, a dense and extremely stiff recycled wood/paper/resin product as a middle layer (with a high Q factor). Following the recommendations of Wayne (WGH) http://www.audiocircle.com/index.php?topic=72808.0, I chose Excel Xpress Gel Polyurethane Adhesive, a semi-flexible expanding polyurethane adhesive as a constrained damping layer between the Richlite and the MDF and to fill in any gaps between the two surfaces, initially fixed in place by hot melt glue, a method he developed. I decided to use ½” SAE F3 felt (with a very low Q factor), to attenuate the low to mid frequencies, glued to the nearly impervious Richlite and also to the exposed MDF with GemTac permanent adhesive http://beaconadhesives.com/cggem-tac.html recommended by Beacon Adhesives. I selected loose Polyester Batting hand-stuffed in the remainder of the enclosure, to further attenuate the mid to lower treble frequencies.

[Tone Depth]

Loudspeaker Enclosure Modification – Part 5:  Fabrication

Wayne assisted me with locating a surplus length of ½” thick Richlite, cutting pieces to size and gluing them to the interior of the MDF walls using Excel Xpress Gel Polyurethane Adhesive, and initially fixed in place using hot melt glue while the polyurethane expanded and cured (see the fourth photo).



Several months later when the summer high temperatures finally abated, I cut to size and glued the ½” SAE F3 felt, to the Richlite layer and also to the remaining exposed MDF using GemTac permanent adhesive (see the fifth, sixth, and seventh photos).







I finally carefully stuffed Polyester Batting in the remainder of the interior to achieve somewhat of a decreasing density gradient from the walls towards the center of the enclosure (see the eighth photo).

[Tone Depth]

Loudspeaker Enclosure Modification – Part 6:  Listening Impressions

My Tangent RS2 speakers now sound much more realistic to me. The bass extends lower, and the midrange has increased clarity. Since the tweeter is sealed, the modifications don’t affect its performance, however, the balance between the treble frequencies and the mid/bass frequencies sounds better matched to me. Using the Odyssey test tracks played on a Redbook CD, a 40Hz tone track played was previously inaudible prior to the modification, and is now clearly audible following the modifications. I read somewhere online that the Audax woofer has a free air resonance of 35Hz; therefore, the modifications have lowered the lower end of the speaker response from 55Hz to between 40Hz and 35Hz. 

[martyo]

Nice job. That's a very significant improvement in low end response. With the increased clarity of the midrange too, must be almost like replacing your speakers. Nice write-up too.

[WGH]

Looks like those speakers are good for another 28 years. Now it's time to upgrade the crossover components. 
Great write up!

Wayne

[planet10]

I read somewhere online that the Audax woofer has a free air resonance of 35Hz; therefore, the modifications have lowered the lower end of the speaker response from 55Hz to between 40Hz and 35Hz. 

Have you actually measured that? I don't recall the RS2s getting to 55 Hz. With a measured system resonance of 70 Hz, response is going to be more like 75 Hz

I don't doubt that you like the bass, but the Richlite is going to decrease the volume of the box, raise the system Q and raise the bass cutoff of the box.

On the other hand the felt (which, being the inside layer) is acting as airspace damping, only the glue holding it on having much to do with the box walls, and the polyfluff will be more effective than the foam at lowering the box Q.

I have fond memories of the Tangents. In particular 1 night when John Greenbank was visiting -- RS4s connected to a Bryston 4B -- he kept promting us turn it up. All of a sudden silence, and acrid white smoke billowing out of the square port on one of the speakers. The resisitors in the XO were underrated and one had got so hot it set the damping foam on fire. 15 min later we had replaced the dead woofer and were "partying" again. At a lower level...

John Greenback had some significant influence on the path of my speaker designs... i would have approached renovating the box more in the vein of the original design philosophy.

dave

[Æ]

Adding damping material is great. While you had the front baffle off, why didn't you add some internal cross braces, bracing? A big dowel or two from one side to the other.

[Æ]

Loudspeaker Enclosure Modification – Part 6:  Listening Impressions

My Tangent RS2 speakers now sound much more realistic to me. The bass extends lower, and the midrange has increased clarity. Since the tweeter is sealed, the modifications don’t affect its performance, however, the balance between the treble frequencies and the mid/bass frequencies sounds better matched to me. Using the Odyssey test tracks played on a Redbook CD, a 40Hz tone track played was previously inaudible prior to the modification, and is now clearly audible following the modifications. I read somewhere online that the Audax woofer has a free air resonance of 35Hz; therefore, the modifications have lowered the lower end of the speaker response from 55Hz to between 40Hz and 35Hz. 

I kind of doubt your modifications gave you deeper bass. More than likely you damped some of the midrange making it seem like you have more bass. To really get any more bass, you would have had to increase the enclosure volume considerably.
You can easily verify system resonance by running a before and after impedance curve. An impedance curve will show you where the system resonance is.

[Tone Depth]

Sorry, I don't have access to any measurement equipment, just referencing the manufacturer's specs of 55 Hz to 32 kHz +/- 3 dB, and listening to test tones to verify the ballpark frequency response.

Yes, adding Richlite decreases the internal volume by a certain percentage, I did the calculation, but no longer recall the figure. It didn't seem to be excessive. Listening after adding the Richlite, as compared with 28 years of listening before adding the Richlite, resulted in audible improvements to the sound quality and clarity, and this includes temporarily reinstalling the foam until I was able to eventually install the felt and batting.

It would be fun and educational to measure the decreased vibrations of the enclosure walls due to adding the Richlite, if I had access to a lab, but since I don't, my qualitative tactile fingertip measurements demonstrated to me the effectiveness of a constrained damping layer and dissimilar Q factors of the MDF and the Richlite. I could feel that the enclosure walls were vibrating much less than before.

My understanding of the attenuation provided by the felt is due to conversion of sound pressure vibrations to heat by friction of the fibers in the felt rubbing together. Yes there is also less surface area directly exposed to the interior sound waves, but there are also plenty of gaps of varying sizes between the felt pads, exposing a lesser area of the MDF surface to the sound waves.

The polyester batting gradation attenuates high frequencies down to a certain point, the felt attenuates mid frequencies further, the Richlite attenuates low frequencies, the constrained damping layer adhesive limits transmission of the low frequencies to the MDF. My understanding of CDL theory is that using dissimilar Q factors, the resulting system Q is further decreased versus being raised as would be the case with rigid laminated layers. My ears confirmed the results of the CDL method, and it's validity especially for this small sealed enclosure. As previously mentioned, hearing an audible 40 Hz test tone, that was inaudible when played prior to the modification, is proof to me.

It may have been posible to install an angled baffle type cross brace, but there isn't a whole lot of real estate inside these small monitors. I suspect that a dowell or rectangular cross-section cross brace would be of limited use, but have never reviewed any data documenting the improvements achieved.

I'd love to run frequency response measurements on my modified speakers, but I have no desire to spend the money for it, it is a needless quantification to me. I'm even more pleased with the sound quality now than before, and that's all that matters to me.

As I initially noted, the basic classic design of the speakers was good in it's era, but somewhat limited. The higher performance retrofit works for me.

I kind of doubt your modifications gave you deeper bass. More than likely you damped some of the midrange making it seem like you have more bass. To really get any more bass, you would have had to increase the enclosure volume considerably.
You can easily verify system resonance by running a before and after impedance curve. An impedance curve will show you where the system resonance is.

[Johnny2Bad]

Aaaah, the RS-2. One of my favourites; although I would not call them "accurate" they did sound sweet and imaged beautifully. About $450 a pair new in the late 70's.
Were it me, I would have left them alone; modern speakers are available if you want more accuracy but they had a certain euphoric magic I found endearing, driven by a Luxman 80W 5L15 Lab Reference Series integrated amp, and a Fidelity Research Fr1mk3 singing on my Denon table. I wish I'd never sold mine.

[Letitroll98]

TD I think perhaps you've eliminated some masking distortions (resonances) in the mid and upper bass that allow you to hear tones that were actually always there.  This ain't a bad thing, evidenced by the improved SQ you noted.  However it's why the suggestions for measurement (perhaps not stated in the most constructive manner) were made.  You can do it pretty cheaply with a Radio Shack SPL meter and the corrections for same put in an excel program.  I've found it both fun and frustrating to do so, but ultimately worthwhile, you might as well if so inclined.

Anyway, I think it's totally kewl that you took this on.  I'm glad you had a good result and posted it here.  Perhaps others more knowledgeable with acoustic suspension designs than I can better answer why you're hearing tones below what you heard previously, even with a reduced enclosure volume.   

[Tone Depth]

Hi '98,

The intent of the modifications were to reduce the speaker backwave frequencies from reflecting forward through the cone, and to reduce the sealed enclosure external resonances. Both of these add undesireable distortion to the music. Improving the efficiency of the internal damping, accomplishes the same effect as using a larger less damped enclosure.

Von Schweikert Audio has conducted the R&D, proven that the technique is effective and currently uses a patented design in their enclosures. Likewise, it is common knowledge that building a larger enclosure extends the low frequency response of a mid/woofer.

I posted this thread to describe my adaptation of the method, and to provide an alternative to others with similar desires to further improve the performance of their speakers. It's all good. 

[planet10]

My understanding of the attenuation provided by the felt is due to conversion of sound pressure vibrations to heat by friction of the fibers in the felt rubbing together. Yes there is also less surface area directly exposed to the interior sound waves, but there are also plenty of gaps of varying sizes between the felt pads,

The energy feed to the cabinet thru the airspace is easily shown to be less than a third of the energy that get sto the cabinet walls. In practise it is less than this.

dave

[Tone Depth]

Hi Dave,

Great comment, thanks!

The energy feed to the cabinet thru the airspace is easily shown to be less than a third of the energy that gets to the cabinet walls. In practise it is less than this.

dave